INDEX

Explanations

expressions of ability or potential outcomes

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

managed to succeed or not

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

RTHOOK

-0.43

ọi

-0.41

****/

-0.39

<()>

-0.39

/******/

-0.38

 insuffisamment

-0.38

ISODE

-0.38

 stran

-0.38

siębior

-0.36

הערות

-0.35

POSITIVE LOGITS

 réussi

0.51

 EClass

0.51

なんとか

0.49

 rius

0.47

Easier

0.44

IntoConstraints

0.44

thâu

0.44

 réus

0.44

何とか

0.44

 riesce

0.43

Activations Density 0.174%

expressions of ability or potential outcomes

managed to succeed or not

No Comments

No Known Activations