INDEX

Explanations

specific structural features and elements within a setting or context

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

specific technical terms and contexts

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

ENSIONS

-0.45

原始内容存档于

-0.43

Nascimento

-0.40

ologically

-0.40

ulaire

-0.40

 fractional

-0.39

sut

-0.39

bairro

-0.38

 extremo

-0.38

திய

-0.37

POSITIVE LOGITS

 дописавши

0.63

twimg

0.58

principalTable

0.50

 AttributeSet

0.49

 tartalomajánló

0.46

postsleuth

0.46

 ExecuteAsync

0.45

 AssemblyTitle

0.45

+#+#

0.45

 Wikimedijinoj

0.43

Activations Density 0.096%

specific structural features and elements within a setting or context

specific technical terms and contexts

No Comments

No Known Activations