INDEX

Explanations

references to copies of documents or records

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Tracce

-0.46

しゃれ

-0.45

tapkan

-0.44

 suun

-0.44

 behaviors

-0.43

 Behavior

-0.42

 BEHAVIOR

-0.42

 moverse

-0.42

 Mathf

-0.42

 Desenvolvimento

-0.41

POSITIVE LOGITS

 copy

0.81

 Copies

0.71

 copies

0.70

Copy

0.70

Copies

0.69

 Copy

0.68

copies

0.65

copy

0.65

 COPY

0.59

 copie

0.56

Activations Density 0.017%

references to copies of documents or records

No Comments

No Known Activations