INDEX

Explanations

words related to different types of actions or activities being done

oai_token-act-pair · gpt-3.5-turbo

the preposition "to" used in various contexts

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

intosh

-0.69

 neighb

-0.68

ask

-0.66

aturday

-0.65

Enlarge

-0.65

Week

-0.64

arna

-0.64

%%

-0.63

iple

-0.62

 swung

-0.62

POSITIVE LOGITS

pless

0.85

 infinity

0.78

 minimize

0.77

 Tycoon

0.75

ilet

0.73

ggles

0.72

 accommodate

0.72

ffee

0.71

 extremes

0.68

 eliminate

0.68

Activations Density 0.070%

words related to different types of actions or activities being done

the preposition "to" used in various contexts

No Known Activations