INDEX

Explanations

occurrences of whitespace or indentation in code or text

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

code and legal text qualifiers

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

 հղումներ

-0.73

الإنجليزية

-0.73

Filmografie

-0.71

 saites

-0.66

 queſta

-0.65

Tikang

-0.65

StoryboardSegue

-0.65

 Infórmanos

-0.64

 stiefe

-0.64

әрмәләр

-0.63

POSITIVE LOGITS

↵↵↵

0.79

↵↵↵↵

0.74

↵↵↵↵↵

0.72

↵↵↵↵↵↵↵

0.63

↵↵↵↵↵↵

0.61

↵↵↵↵↵↵↵↵

0.60

↵↵↵↵↵↵↵↵↵

0.58

↵↵↵↵↵↵↵↵↵↵↵

0.56

↵↵↵↵↵↵↵↵↵↵

0.52

uxxxx

0.50

Activations Density 0.010%

occurrences of whitespace or indentation in code or text

code and legal text qualifiers

No Comments

No Known Activations