INDEX

Explanations

arguments and discussions for and against specific policies or laws

oai_token-act-pair · gpt-3.5-turbo

terms related to opposition and criticism in discussions about policies or measures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

owship

-0.66

CLASSIFIED

-0.64

eret

-0.60

Redd

-0.60

 correctional

-0.59

Coach

-0.59

 curfew

-0.58

Located

-0.57

 Abyss

-0.57

versible

-0.57

POSITIVE LOGITS

 argue

1.11

 allege

1.08

 contend

1.07

 accuse

1.04

ervatives

1.03

 argued

0.97

paces

0.95

say

0.95

 concede

0.94

 insist

0.89

Activations Density 0.071%

arguments and discussions for and against specific policies or laws

terms related to opposition and criticism in discussions about policies or measures

No Comments

No Known Activations