INDEX

Explanations

programming constructs related to code structure and syntax

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

code identifiers and punctuation

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

Personendaten

-0.85

出版年

-0.84

 виправивши

-0.79

 Infórmanos

-0.76

 ویکی‌پدی

-0.74

Cyfeiriadau

-0.73

DoubleQuotes

-0.71

 HFILL

-0.70

Datuak

-0.69

səhifə

-0.68

POSITIVE LOGITS

LookAnd

0.50

XmlAccessType

0.44

struct

0.33

cstdlib

0.31

 Folgen

0.31

NamedQueries

0.29

FDRE

0.29

tonode

0.28

 Zust

0.28

 Tropfen

0.28

Activations Density 0.017%

programming constructs related to code structure and syntax

code identifiers and punctuation

No Comments

No Known Activations