INDEX

Explanations

references to specific entities or figures in discussions about policies or social issues

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

founded by, we had

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

↹

-0.39

дент

-0.35

lesias

-0.35

 Barlow

-0.34

getFullYear

-0.34

mat

-0.32

 Fiske

-0.32

<h1>

-0.31

ună

-0.31

tre

-0.30

POSITIVE LOGITS

ValueStyle

0.67

 BoxFit

0.63

الدراسه

0.63

ьаж

0.61

DockStyle

0.60

ViewImports

0.60

RenderAtEndOf

0.58

awtextra

0.56

 intptr

0.56

 Roskov

0.56

Activations Density 0.440%

references to specific entities or figures in discussions about policies or social issues

founded by, we had

No Comments

No Known Activations