INDEX

Explanations

phrases that indicate temporal or numerical details related to events or statistics

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

Followed by time units

np_acts-logits-general · gemini-2.0-flash

units of time

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

 Held

-0.43

 prochain

-0.42

last

-0.40

 proto

-0.40

ef

-0.39

 certo

-0.39

times

-0.38

 שוליים

-0.38

miştir

-0.37

 való

-0.37

POSITIVE LOGITS

 season

1.25

 month

1.07

 year

1.05

AddTagHelper

1.02

season

0.98

 week

0.94

 semester

0.93

 Season

0.90

 SEASON

0.89

Season

0.88

Activations Density 0.131%

phrases that indicate temporal or numerical details related to events or statistics

Followed by time units

units of time

No Comments

No Known Activations