INDEX

Explanations

words related to objects, particularly those associated with creation, installation, or modifications

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

programming and technical terms

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

 nakalista

-0.40

fä

-0.36

нав

-0.33

Elect

-0.33

 furthest

-0.31

entliche

-0.31

Insn

-0.31

 elect

-0.31

 specific

-0.31

 everything

-0.30

POSITIVE LOGITS

DockStyle

0.86

MLLoader

0.59

tagHelper

0.59

 FetchType

0.55

 gesteld

0.52

̈́

0.52

 BoxDecoration

0.52

 ProtoMessage

0.51

WEBPACK

0.50

(!__

0.50

Activations Density 0.044%

words related to objects, particularly those associated with creation, installation, or modifications

programming and technical terms

No Comments

No Known Activations