INDEX

Explanations

references to edges in graph structures or related concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_0/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 ciudadana

-0.71

 cielos

-0.67

 Absicht

-0.64

 avoient

-0.63

 démocr

-0.61

 fédé

-0.59

 espoir

-0.59

 pouvoit

-0.59

 feroit

-0.58

 Italij

-0.58

POSITIVE LOGITS

ніципалі

0.71

 Edge

0.65

BeginContext

0.64

AndEndTag

0.61

Lex

0.60

Cro

0.59

 EDGE

0.59

 но

0.59

 Но

0.58

分

0.57

Activations Density 0.344%

references to edges in graph structures or related concepts

No Comments

No Known Activations