INDEX

Explanations

phrases indicating statements or quotations from individuals, particularly in a news context

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

capitalized words after 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

MessageTagHelper

-0.61

RenderAtEndOf

-0.54

 ब्रेकडाउन

-0.53

دانشنامهٔ

-0.52

 emojis

-0.52

 COVID

-0.51

 للاسماء

-0.51

 https

-0.51

 GoFundMe

-0.51

 resourceCulture

-0.50

POSITIVE LOGITS

RegressionTest

0.47

chakra

0.42

BeginContext

0.36

struts

0.36

HttpRequest

0.36

 aDecoder

0.34

Shock

0.34

tira

0.34

 reporter

0.33

shock

0.32

Activations Density 0.089%

phrases indicating statements or quotations from individuals, particularly in a news context

capitalized words after punctuation

No Comments

No Known Activations

phrases indicating statements or quotations from individuals, particularly in a news context

capitalized words after punctuation

No Comments

No Known Activations