INDEX

Explanations

expressions of dialogue or interactions between characters

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

actions and their consequences

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

-0.36

achten

-0.33

 durer

-0.32

 ache

-0.32

izo

-0.31

 funding

-0.31

or

-0.30

userDao

-0.30

 precios

-0.30

 fund

-0.30

POSITIVE LOGITS

jsii

0.72

'{@

0.71

 betweenstory

0.70

новниш

0.64

enterOuterAlt

0.63

UrlResolution

0.63

">//

0.63

AndEndTag

0.62

 surla

0.61

 propOrder

0.61

Activations Density 0.039%

expressions of dialogue or interactions between characters

actions and their consequences

No Comments

No Known Activations