INDEX

Explanations

numerical values and codes

oai_token-act-pair · gpt-3.5-turbo

numerical data and statistics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 nomine

-0.96

annis

-0.86

 contrace

-0.79

 corrid

-0.78

 neighb

-0.74

 referen

-0.73

 hemor

-0.71

 misunder

-0.70

 interf

-0.69

ModLoader

-0.69

POSITIVE LOGITS

 partName

0.93

df

0.81

 attRot

0.77

0.73

dict

0.72

0.71

0.70

çļĦ

0.70

fc

0.68

Activations Density 0.122%

numerical values and codes

numerical data and statistics

No Comments

No Known Activations