INDEX

Explanations

words related to police operations or investigations

oai_token-act-pair · gpt-3.5-turbo

references to sting operations

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

ufact

-0.88

iasco

-0.72

ACTED

-0.70

NCT

-0.65

ocally

-0.63

 Supplement

-0.63

unicip

-0.62

 Colleg

-0.62

ordan

-0.61

udeau

-0.61

POSITIVE LOGITS

 sting

1.19

rays

1.14

ray

1.14

ega

0.95

ingly

0.94

lers

0.86

iest

0.78

Ray

0.78

iness

0.76

ofi

0.76

Activations Density 0.017%

words related to police operations or investigations

references to sting operations

No Comments

No Known Activations