INDEX

Explanations

names and references to specific programming or technical terms

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

auto-generated code

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

 Biôgrafia

-0.62

 AssemblyProduct

-0.57

NameInMap

-0.49

UserScript

-0.49

 ejerc

-0.48

 Савезне

-0.48

Rankings

-0.47

 tričko

-0.47

perity

-0.47

Success

-0.47

POSITIVE LOGITS

awaiter

1.15

oneofs

0.45

__*/

0.45

DebuggerStep

0.40

AndEndTag

0.39

Kaynakça

0.36

 resourceCulture

0.36

GTCX

0.35

errHandler

0.34

Viited

0.33

Activations Density 0.000%

names and references to specific programming or technical terms

auto-generated code

No Comments

No Known Activations

names and references to specific programming or technical terms

auto-generated code

No Comments

No Known Activations