INDEX

Explanations

proper nouns associated with prestigious awards

oai_token-act-pair · gpt-3.5-turbo

references to various awards

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

oidal

-0.75

cos

-0.72

CY

-0.69

vironments

-0.68

apse

-0.66

 breathing

-0.66

ths

-0.65

 inconven

-0.64

folk

-0.64

 troubles

-0.63

POSITIVE LOGITS

 Award

1.01

Winner

0.99

 awarded

0.96

 Winners

0.95

 Awards

0.89

 Trophy

0.87

 Winning

0.86

 Prize

0.86

 Winner

0.85

eous

0.85

Activations Density 0.021%

proper nouns associated with prestigious awards

references to various awards

No Comments

No Known Activations