INDEX

Explanations

phrases that contain names, identifiers, or references to structured data formats

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

 بيها

-0.78

NameInMap

-0.72

AnchorStyles

-0.68

writeField

-0.68

')")

-0.68

RUnlock

-0.67

 propOrder

-0.66

ValueStyle

-0.65

قایناقلار

-0.65

CppCodeGen

-0.63

POSITIVE LOGITS

XYZ

0.88

XYZ

0.86

xyz

0.76

xyz

0.75

ABC

0.71

XXX

0.71

XXX

0.67

XX

0.65

XX

0.63

ABC

0.63

Activations Density 0.769%

phrases that contain names, identifiers, or references to structured data formats

No Comments

No Known Activations