INDEX

Explanations

punctuation marks and whitespace-related formatting

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

EDEFAULT

-0.74

RenderAtEndOf

-0.73

 ModelExpression

-0.73

 AssemblyCulture

-0.72

MessageOf

-0.67

principalTable

-0.67

<pad>

-0.65

 Bioaccumulative

-0.65

<unused8>

-0.65

<unused28>

-0.65

POSITIVE LOGITS

import

0.40

.*;

0.35

'}';

0.34

Zi

0.31

.*;

0.30

 comté

0.30

 AutoMapper

0.29

Slf

0.29

Activations Density 0.000%

punctuation marks and whitespace-related formatting

No Comments

No Known Activations