INDEX

Explanations

words related to fraudulent actions or schemes

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "sw" in various contexts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

pora

-0.80

 burying

-0.65

 negligent

-0.65

_-

-0.64

ICAL

-0.63

NESS

-0.63

 withholding

-0.63

 paving

-0.63

 cigarettes

-0.62

oglu

-0.62

POSITIVE LOGITS

aths

0.92

sw

0.92

atches

0.90

arf

0.88

immers

0.88

athed

0.87

addle

0.85

apon

0.85

erd

0.84

anky

0.83

Activations Density 0.004%

words related to fraudulent actions or schemes

occurrences of the word "sw" in various contexts

No Comments

No Known Activations