INDEX

Explanations

punctuation, specifically semicolons

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

<start_of_turn> user

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

Negative Logits

 betweenstory

-0.82

sizeCache

-0.77

SequentialGroup

-0.76

IsMutable

-0.74

writeFieldEnd

-0.72

makeConstraints

-0.71

 ModelExpression

-0.69

новниш

-0.68

InSection

-0.68

 jspb

-0.68

POSITIVE LOGITS

↵↵

0.63

↵↵↵

0.42

class

0.42

 Please

0.41

↵

0.39

 Come

0.36

namespace

0.36

linge

0.34

 class

0.33

/**

0.33

Activations Density 0.000%

punctuation, specifically semicolons

<start_of_turn> user

No Known Activations