INDEX

Explanations

mathematical variables and references to induction in proofs

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

mathematical and code contexts

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

얻

-0.35

 diting

-0.33

-0.32

und

-0.29

 Rumuni

-0.29

and

-0.29

mybatisplus

-0.29

-0.28

al

-0.28

op

-0.28

POSITIVE LOGITS

AndEndTag

1.21

RenderAtEndOf

0.96

:✨

0.91

SharedCtor

0.91

 CreateTagHelper

0.82

 Wikimedijinoj

0.82

UrlResolution

0.82

SequentialGroup

0.82

LookAnd

0.81

TagMode

0.79

Activations Density 0.030%

mathematical variables and references to induction in proofs

mathematical and code contexts

No Comments

No Known Activations