INDEX

Explanations

punctuation marks and method calls in code excerpts

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

method calls and programming elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

{},

-0.47

StructEnd

-0.47

 nessun

-0.46

TypedDataSet

-0.44

/$',

-0.42

 sumpay

-0.42

 cerâmica

-0.42

toContain

-0.41

]+$

-0.41

/**

-0.41

POSITIVE LOGITS

 GoogleFonts

0.55

 bandeira

0.46

ImageContext

0.45

sieurs

0.44

blica

0.44

ulemon

0.43

Betyg

0.42

 consultato

0.42

izoen

0.41

んですよね

0.41

Activations Density 0.016%

punctuation marks and method calls in code excerpts

method calls and programming elements

No Comments

No Known Activations