INDEX

Explanations

adverbs and descriptive words

oai_token-act-pair · gpt-3.5-turbo

expressions of disbelief or criticism

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

successfully

-0.71

interrupted

-0.62

usalem

-0.62

 secured

-0.61

 respectively

-0.61

 colon

-0.60

 benefic

-0.59

 predetermined

-0.59

kamp

-0.59

 grievance

-0.58

POSITIVE LOGITS

atches

0.80

ÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤ

0.73

eps

0.70

Ĥİ

0.69

 compared

0.67

than

0.67

deals

0.66

til

0.65

notations

0.65

heres

0.64

Activations Density 0.405%

adverbs and descriptive words

expressions of disbelief or criticism

No Comments

No Known Activations