INDEX

Explanations

phrases related to a specific organization or group

oai_token-act-pair · gpt-3.5-turbo

references to groups and organizations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

shire

-0.71

cale

-0.68

èª

-0.64

Rah

-0.62

 concess

-0.62

LEASE

-0.61

 Latest

-0.60

yy

-0.60

lig

-0.60

 Shard

-0.59

POSITIVE LOGITS

think

0.96

members

0.95

leader

0.94

arettes

0.92

arette

0.91

 disbanded

0.90

roups

0.87

 members

0.87

rooms

0.82

mates

0.80

Activations Density 0.050%

phrases related to a specific organization or group

references to groups and organizations

No Comments

No Known Activations