INDEX

Explanations

elements related to programming or code structures

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

code and math syntax

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

évaluateur

-0.57

tvguidetime

-0.56

 Taktlose

-0.56

Personendaten

-0.53

ArrowToggle

-0.47

SPATH

-0.47

сылкі

-0.46



-0.45

EndProject

-0.45

 naman

-0.44

POSITIVE LOGITS

 <>",

0.46

MessageOf

0.43

 zeker

0.42

 הבי

0.41

im

0.40

 wskaz

0.40

 chimiques

0.40

 separado

0.40

 independiente

0.39

 zimowa

0.38

Activations Density 0.004%

elements related to programming or code structures

code and math syntax

No Comments

No Known Activations