INDEX

Explanations

instances of impressive achievements or remarkable accomplishments

oai_token-act-pair · gpt-3.5-turbo

references to accomplishments or notable achievements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 contracting

-0.73

srf

-0.72

iant

-0.71

Dot

-0.69

 consultation

-0.67

 Galile

-0.66

iod

-0.65

 Leilan

-0.65

holm

-0.63

 Quarterly

-0.63

POSITIVE LOGITS

hered

1.77

 feat

1.02

 feats

0.90

hest

0.85

chery

0.84

reys

0.83

icles

0.83

ree

0.82

chers

0.81

gerald

0.79

Activations Density 0.010%

instances of impressive achievements or remarkable accomplishments

references to accomplishments or notable achievements

No Comments

No Known Activations