INDEX

Explanations

references to specific numerical values related to evaluations or performance metrics

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

numbers followed by 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

migrationBuilder

-0.49

 klok

-0.46

 LeBlanc

-0.46

✴

-0.45

 Horowitz

-0.45

 Branch

-0.45

 BRANCH

-0.45

 Brock

-0.44

 Moff

-0.44

 Microb

-0.43

POSITIVE LOGITS

 getUserId

0.57

 ConstraintSet

0.56

 AssertionError

0.53

 Parton

0.49

 sqlSession

0.49

icestershire

0.48

ాయి

0.48

 didst

0.48

 spé

0.48

 JSTOR

0.48

Activations Density 0.207%

references to specific numerical values related to evaluations or performance metrics

numbers followed by punctuation

No Comments

No Known Activations