INDEX

Explanations

proper nouns related to characters or locations

oai_token-act-pair · gpt-3.5-turbo

the name "Howard."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 thirst

-0.69

pir

-0.66

dil

-0.66

 boxed

-0.63

 sequels

-0.61

 provoking

-0.61

 elev

-0.60

adv

-0.59

aud

-0.59

rh

-0.59

POSITIVE LOGITS

udeau

0.83

icago

0.82

soType

0.80

Ont

0.80

Ohio

0.77

apolis

0.77

OME

0.75

NET

0.74

Ò

0.73

Via

0.72

Activations Density 0.000%

proper nouns related to characters or locations

the name "Howard."

No Comments

No Known Activations