INDEX

Explanations

sequence of commands or operations typically found in programming or coding contexts

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

code, configuration, and test cases

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

parsedMessage

-0.81

 disambiguazione

-0.72

httphttps

-0.72

IndentedString

-0.65

DockStyle

-0.65

IsContent

-0.63

 समीक्षाओं

-0.62

rrggbb

-0.61

complexType

-0.59

isContained

-0.59

POSITIVE LOGITS

<strong>

0.42

asInstanceOf

0.40

FIX

0.40

台

0.38

Fix

0.38

howto

0.38

Winter

0.38

 caixa

0.38

Зна

0.38

Support

0.36

Activations Density 0.036%

sequence of commands or operations typically found in programming or coding contexts

code, configuration, and test cases

No Comments

No Known Activations