INDEX

Explanations

instances of exceptions or exclusions

oai_token-act-pair · gpt-3.5-turbo

phrases expressing conditional or qualifying contexts

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

rift

-0.79

orthy

-0.78

ongyang

-0.76

yssey

-0.72

kefeller

-0.72

orian

-0.72

ahead

-0.68

 Cosponsors

-0.67

oured

-0.67

cean

-0.66

POSITIVE LOGITS

 occasional

0.83

 exceptions

0.80

 caveats

0.74

 caveat

0.68

 occasionally

0.67

 ones

0.66

 Exception

0.66

emort

0.65

 glaring

0.64

 faint

0.63

Activations Density 0.093%

instances of exceptions or exclusions

phrases expressing conditional or qualifying contexts

No Comments

No Known Activations