INDEX

Explanations

technical or legal terms related to processes and regulations

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

language tags; technical terms; code 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

principalColumn

-0.51

 recuer

-0.40

 contigo

-0.39

StoryboardSegue

-0.38

 mengen

-0.38

 deltag

-0.38

poj

-0.37

ẠI

-0.36

zugehen

-0.36

 interesan

-0.35

POSITIVE LOGITS

parsedMessage

0.72

endphp

0.61

 surla

0.57

 betweenstory

0.55

__*/

0.55

:");

0.53

WEBPACK

0.52

"];

0.51

expandindo

0.51

 otomatig

0.51

Activations Density 0.028%

technical or legal terms related to processes and regulations

language tags; technical terms; code syntax

No Comments

No Known Activations