INDEX

Explanations

references to time durations and intervals

oai_token-act-pair · gpt-4o-mini Triggered by @bot

time durations, especially minutes, with high activation for phrases like "5 minutes" and "few minutes".

oai_token-act-pair · claude-3-7-sonnet-20250219 Triggered by @neilrathi

mentions of quantified short time intervals (durations) and nearby punctuation or formatting such as hyphens, commas, and periods.

oai_token-act-pair · gpt-5 Triggered by @jyhe0408

Looking at the activations, this neuron strongly activates on the word "minute" or "minutes" (values ranging from ~57-67), particularly when it appears in contexts describing durations or time periods. The neuron is detecting references to time durations measured in minutes.

oai_token-act-pair · claude-4-5-sonnet Triggered by @jyhe0408

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 koszulka

-0.44

 Weing

-0.43

 anzunehmen

-0.43

 elä

-0.42

veloppement

-0.41

 Públicas

-0.40

 çift

-0.40

jenigen

-0.40

IsMutable

-0.40

spesies

-0.40

POSITIVE LOGITS

Minute

1.17

 Minute

1.17

 minute

1.12

minute

1.09

 minutes

1.08

getMinutes

1.07

minutes

1.05

 MINUTE

1.03

Minutes

0.97

 minuto

0.97

Activations Density 0.078%

references to time durations and intervals

time durations, especially minutes, with high activation for phrases like "5 minutes" and "few minutes".

mentions of quantified short time intervals (durations) and nearby punctuation or formatting such as hyphens, commas, and periods.

Looking at the activations, this neuron strongly activates on the word "minute" or "minutes" (values ranging from ~57-67), particularly when it appears in contexts describing durations or time periods. The neuron is detecting references to time durations measured in minutes.

No Comments

No Known Activations

references to time durations and intervals

time durations, especially minutes, with high activation for phrases like "5 minutes" and "few minutes".

mentions of quantified short time intervals (durations) and nearby punctuation or formatting such as hyphens, commas, and periods.

Looking at the activations, this neuron strongly activates on the word "minute" or "minutes" (values ranging from ~57-67), particularly when it appears in contexts describing durations or time periods. The neuron is detecting references to time durations measured in minutes.

No Comments

No Known Activations

references to time durations and intervals

time durations, especially minutes, with high activation for phrases like "5 minutes" and "few minutes".

mentions of quantified short time intervals (durations) and nearby punctuation or formatting such as hyphens, commas, and periods.

Looking at the activations, this neuron strongly activates on the word "minute" or "minutes" (values ranging from ~57-67), particularly when it appears in contexts describing durations or time periods. The neuron is detecting **references to time durations measured in minutes**.

No Comments

No Known Activations

references to time durations and intervals

time durations, especially minutes, with high activation for phrases like "5 minutes" and "few minutes".

mentions of quantified short time intervals (durations) and nearby punctuation or formatting such as hyphens, commas, and periods.

Looking at the activations, this neuron strongly activates on the word "minute" or "minutes" (values ranging from ~57-67), particularly when it appears in contexts describing durations or time periods. The neuron is detecting **references to time durations measured in minutes**.

No Comments

No Known Activations

Looking at the activations, this neuron strongly activates on the word "minute" or "minutes" (values ranging from ~57-67), particularly when it appears in contexts describing durations or time periods. The neuron is detecting references to time durations measured in minutes.

Looking at the activations, this neuron strongly activates on the word "minute" or "minutes" (values ranging from ~57-67), particularly when it appears in contexts describing durations or time periods. The neuron is detecting references to time durations measured in minutes.