INDEX

Explanations

mentions related to specific styles or aspects of style

oai_token-act-pair · gpt-3.5-turbo

references to various styles in different contexts

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

orb

-0.76

MS

-0.70

enary

-0.70

bley

-0.70

ajor

-0.68

VEN

-0.68

riel

-0.67

eor

-0.67

ertodd

-0.67

ora

-0.66

POSITIVE LOGITS

heet

0.87

ologies

0.87

 sheets

0.83

 sheet

0.80

 cues

0.78

 preference

0.78

 guides

0.76

 guide

0.75

face

0.75

 guideline

0.71

Activations Density 0.048%

mentions related to specific styles or aspects of style

references to various styles in different contexts

No Known Activations