INDEX

Explanations

contractions or contractions with punctuation in them

oai_token-act-pair · gpt-3.5-turbo

negations or phrases indicating uncertainty

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

 pione

-0.67

ò

-0.62

 exting

-0.62

redients

-0.61

ayer

-0.59

ñ

-0.59

VERTISEMENT

-0.59

Þ

-0.58

 subur

-0.58

idable

-0.58

POSITIVE LOGITS

 myself

1.06

 personally

0.97

 hesitate

0.95

 believe

0.94

 doubt

0.93

 condone

0.92

 recommend

0.90

 know

0.86

 think

0.85

 sure

0.84

Activations Density 0.131%

contractions or contractions with punctuation in them

negations or phrases indicating uncertainty

No Comments

No Known Activations