INDEX

Explanations

mentions of things being the same or identical

oai_token-act-pair · gpt-3.5-turbo

references to the concept of matching

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Ïī

-0.70

OLOG

-0.70

********************************

-0.69

ase

-0.67

Daily

-0.66

################################

-0.65

ember

-0.65

aug

-0.65

gom

-0.64

bra

-0.64

POSITIVE LOGITS

 matching

1.10

 matched

1.05

 matches

0.95

ees

0.81

 match

0.80

inances

0.75

 mism

0.74

paren

0.73

poons

0.71

 pairs

0.71

Activations Density 0.010%

mentions of things being the same or identical

references to the concept of matching

No Comments

No Known Activations