INDEX

Explanations

pronouns 'it' with various associated descriptions or statements

oai_token-act-pair · gpt-3.5-turbo

instances of text termination or empty sections

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 Credit

-0.53

 Eighth

-0.52

 Goldstein

-0.50

 Polk

-0.49

 Pell

-0.49

————————

-0.49

 Jensen

-0.48

 absentee

-0.48

 Union

-0.48

Pav

-0.48

POSITIVE LOGITS

alian

1.13

unes

1.04

chy

0.97

 doesnt

0.96

 depends

0.94

ain

0.94

anium

0.93

zbollah

0.87

 hurts

0.87

 rains

0.84

Activations Density 0.051%

pronouns 'it' with various associated descriptions or statements

instances of text termination or empty sections

No Comments

No Known Activations