INDEX

Explanations

punctuation marks and numerical values

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

 ainfi

-0.74

MLLoader

-0.71

 arbete

-0.69

 AppColors

-0.66

 ſever

-0.66

 pleaſure

-0.65

 feroit

-0.65

 raiſ

-0.65

 auffi

-0.64

الدراسه

-0.64

POSITIVE LOGITS

enumi

0.77

guy

0.61

Clik

0.60

getDoctrine

0.60

 kind

0.59

 guys

0.59

↵↵

0.57

BeginContext

0.54

 pretty

0.54

0.53

Activations Density 0.001%

punctuation marks and numerical values

No Comments

No Known Activations