INDEX

Explanations

legal and investigative terms related to cause determination

oai_token-act-pair · gpt-3.5-turbo

references to causes or reasons behind events or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Seasons

-0.70

illet

-0.70

 Sachs

-0.66

 Storm

-0.64

PDATE

-0.62

Ku

-0.61

poons

-0.60

atari

-0.60

 Tips

-0.60

ilings

-0.60

POSITIVE LOGITS

 cele

1.43

way

0.92

ality

0.84

 thereof

0.76

cause

0.76

forge

0.75

 celeb

0.73

vier

0.70

ways

0.69

conom

0.68

Activations Density 0.042%

legal and investigative terms related to cause determination

references to causes or reasons behind events or conditions

No Comments

No Known Activations