INDEX

Explanations

mathematical operations and expressions, particularly involving parentheses and various arithmetic symbols

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

mathematical expressions with parentheses

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

 argint

-0.64

+#+#

-0.62

PreInfinity

-0.59

 ComVisible

-0.58

yntaxException

-0.57

⟬

-0.57

<unused8>

-0.57

[@BOS@]

-0.57

<unused3>

-0.57

<unused14>

-0.57

POSITIVE LOGITS

帖最后由

0.42

concat

0.36

 fuse

0.32

hey

0.31

发表于

0.31

pen

0.31

0.30

 dare

0.30

se

0.30

Activations Density 0.263%

mathematical operations and expressions, particularly involving parentheses and various arithmetic symbols

mathematical expressions with parentheses

No Comments

No Known Activations