INDEX

Explanations

medical conditions characterized by pathology or political keywords

oai_token-act-pair · gpt-3.5-turbo

terms related to auditory phenomena or assessments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

uggle

-0.68

topia

-0.68

 Brit

-0.63

 rulings

-0.61

boat

-0.61

 Occupations

-0.61

rieved

-0.60

 clearance

-0.60

 doct

-0.59

 approvals

-0.58

POSITIVE LOGITS

phas

0.96

etic

0.92

ized

0.92

icum

0.91

ciation

0.87

etically

0.87

isc

0.87

izes

0.86

ization

0.82

etics

0.81

Activations Density 0.020%

medical conditions characterized by pathology or political keywords

terms related to auditory phenomena or assessments

No Comments

No Known Activations