INDEX

Explanations

programming syntax related to control structures and loops

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

code with while loops

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

 zwiſchen

-0.65

脚注の使い方

-0.61

 MainAxisSize

-0.57

exitRule

-0.56

 հղումներ

-0.54

<unused74>

-0.54

<unused3>

-0.54

<unused16>

-0.53

<unused8>

-0.53

[@BOS@]

-0.53

POSITIVE LOGITS

 NSCoder

0.59

 suivante

0.35

UserScript

0.31

moveToNext

0.31

ModelSerializer

0.30

complexContent

0.30

ientôt

0.30

 prø

0.29

jęt

0.29

try

0.29

Activations Density 0.004%

programming syntax related to control structures and loops

code with while loops

No Comments

No Known Activations