INDEX

Explanations

phrases indicating personal opinions or feelings about situations

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

RegressionTest

-0.65

躇

-0.65

vábbi

-0.62

andaag

-0.59

stranded

-0.58

כז

-0.56

MERCE

-0.55

stdc

-0.55

 invokingState

-0.55

揄

-0.55

POSITIVE LOGITS

 does

1.65

did

1.65

DID

1.48

do

1.43

 DOES

1.36

did

1.36

does

1.32

sí

1.31

确实

1.28

 כן

1.25

Activations Density 0.340%

phrases indicating personal opinions or feelings about situations

No Comments

No Known Activations