INDEX

Explanations

terms related to varied subjects such as medical conditions, movies, community, and others without a clear specific focus

oai_token-act-pair · gpt-3.5-turbo

references to criminal activities and legal implications

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

anwhile

-0.60

ĪĴ

-0.57

pires

-0.56

ousands

-0.56

)"

-0.53

actionDate

-0.53

 trillions

-0.52

aturdays

-0.51

illions

-0.51

usalem

-0.50

POSITIVE LOGITS

0.55

 arrang

0.53

 predicament

0.52

 whereby

0.51

 shenan

0.50

yssey

0.50

 experien

0.50

 behaviour

0.49

 wherein

0.48

 strategy

0.48

Activations Density 1.394%

terms related to varied subjects such as medical conditions, movies, community, and others without a clear specific focus

references to criminal activities and legal implications

No Comments

No Known Activations