INDEX

Explanations

headings or start structures in the text

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

principalColumn

-0.80

 propOrder

-0.78

 himo

-0.74

InitVars

-0.73

SharedDtor

-0.70

NoSuchAlgorithm

-0.69

PropertyChanging

-0.68

enderror

-0.68

WriteBarrier

-0.68

onAttach

-0.68

POSITIVE LOGITS

 traceable

0.48

🇫

0.47

 recycled

0.46

Lorem

0.44

 fund

0.43

 funds

0.42

водства

0.42

labels

0.41

 recuperação

0.41

RunWith

0.41

Activations Density 0.009%

headings or start structures in the text

No Known Activations