INDEX

Explanations

configuration and default settings-related terms

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

default or fallback

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

 виправивши

-0.42

AccessorTable

-0.40

quelize

-0.36

chaffenheit

-0.35

mac

-0.34

IUrlHelper

-0.33

Cal

-0.33

 dismantled

-0.32

følgelig

-0.32

moveToFirst

-0.32

POSITIVE LOGITS

 unknownFields

0.58

ImageContext

0.53

WithIOException

0.48

 defaultValue

0.48

featureID

0.46

AnchorTagHelper

0.46

aarrggbb

0.45

 initComponents

0.45

RTLR

0.45

writeFieldEnd

0.45

Activations Density 0.009%

configuration and default settings-related terms

default or fallback

No Comments

No Known Activations