INDEX

Explanations

phrases indicating location and context within a discussion

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

in

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

 defaultstate

-0.40

bootstrapcdn

-0.38

ついでに

-0.36

何が

-0.36

 antaranya

-0.35

via

-0.34

Various

-0.34

 various

-0.34

RTLI

-0.34

めっちゃ

-0.33

POSITIVE LOGITS

AndEndTag

0.74

 nowadays

0.68

IntoConstraints

0.66

Diwedd

0.63

 today

0.56

 AssemblyCulture

0.55

 Roskov

0.54

目前的

0.54

 ffilmiau

0.54

Nowadays

0.54

Activations Density 0.139%

phrases indicating location and context within a discussion

in

No Comments

No Known Activations