INDEX

Explanations

file path or directory-related terms

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

working directory or weight

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

 piú

-0.67

 Infórmanos

-0.59

odeon

-0.59

\{\\

-0.59

ValueStyle

-0.57

 snippetHide

-0.57

>${

-0.57

oncé

-0.56

verwijspagina

-0.56

tvguidetime

-0.55

POSITIVE LOGITS

pwd

1.02

pwd

0.93

Pwd

0.92

WD

0.79

wd

0.79

PWD

0.69

WD

0.65

cwd

0.63

cwd

0.62

WT

0.61

Activations Density 0.012%

file path or directory-related terms

working directory or weight

No Comments

No Known Activations