INDEX

Explanations

programming-related error handling and logging statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

fjspx

-0.74

 sota

-0.61

########.

-0.60

 שוליים

-0.54

 bağlantılar

-0.52

 nahilalakip

-0.52

mybatisplus

-0.51

 cutting

-0.50

mable

-0.50

 persa

-0.50

POSITIVE LOGITS

gjenge

0.55

 rând

0.53

UIAlert

0.53

runOnUiThread

0.52

 UIAlert

0.52

 näytte

0.51

setPixel

0.50

Hozzáférés

0.49

AndEndTag

0.49

cauc

0.48

Activations Density 0.006%

programming-related error handling and logging statements

No Comments

No Known Activations