INDEX

Explanations

statements presenting results or evidence

oai_token-act-pair · gpt-3.5-turbo

assertive statements about evidence and results

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

emis

-0.75

thood

-0.71

ciation

-0.67

[|

-0.67

iage

-0.67

anism

-0.66

onomous

-0.66

isites

-0.64

outhern

-0.64

vention

-0.63

POSITIVE LOGITS

 damning

1.25

 conclusive

1.18

 unequiv

1.13

 anecdotal

1.09

 incon

1.03

 alarming

1.03

 conflicting

1.00

 startling

0.96

 reassuring

0.96

 discouraging

0.93

Activations Density 0.150%

statements presenting results or evidence

assertive statements about evidence and results

No Comments

No Known Activations