INDEX

Explanations

programming language keywords and syntax elements

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

share or namespace prefixes

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

 resourceCulture

-0.61

تقاوى

-0.57

IsContent

-0.52

 pleaſure

-0.52

EDEFAULT

-0.51

 juſ

-0.47

Diwedd

-0.46

 beſt

-0.46

 ſtand

-0.45

 raiſ

-0.44

POSITIVE LOGITS

msgTypes

0.46

ectoria

0.43

UMNS

0.40

astrous

0.40

 reviving

0.38

 Revival

0.38

 waving

0.37

setViewportView

0.37

 revival

0.37

Ast

0.37

Activations Density 0.008%

programming language keywords and syntax elements

share or namespace prefixes

No Comments

No Known Activations