INDEX

Explanations

phrases indicating quantities and degrees of experience or opinion

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 38-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_38/width_16k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.38.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

時に

-0.35

pholes

-0.35

 lest

-0.34

tagHelperRunner

-0.34

 willingly

-0.34

 indisponible

-0.34

marginHorizontal

-0.34

 today

-0.34

uber

-0.34

 alongside

-0.33

POSITIVE LOGITS

 respects

1.23

 respect

1.02

 regards

0.94

respect

0.88

 regard

0.85

 instances

0.84

Respect

0.82

 sentidos

0.80

 ways

0.76

 senses

0.74

Activations Density 0.393%

phrases indicating quantities and degrees of experience or opinion

No Comments

No Known Activations

phrases indicating quantities and degrees of experience or opinion

No Comments

No Known Activations