INDEX

Explanations

instances of coercion or being compelled to do something against one's will

oai_token-act-pair · gpt-3.5-turbo

instances of the word "forced" indicating coercion or compulsion

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Offense

-0.60

Enh

-0.59

ership

-0.58

izoph

-0.57

 Lights

-0.56

pport

-0.54

 Prediction

-0.54

iosyncr

-0.54

aird

-0.53

etheless

-0.53

POSITIVE LOGITS

 into

1.12

 thereto

0.91

to

0.87

 onto

0.87

 INTO

0.83

 awake

0.83

otom

0.82

into

0.81

 aback

0.81

tto

0.75

Activations Density 0.056%

instances of coercion or being compelled to do something against one's will

instances of the word "forced" indicating coercion or compulsion

No Comments

No Known Activations