INDEX

Explanations

strong statements about standing up for justice, accountability, and equality

oai_token-act-pair · gpt-3.5-turbo

negative phrases indicating resistance or refusal toward actions and behaviors

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

umably

-0.74

lycer

-0.74

avorite

-0.72

Ear

-0.68

ilar

-0.68

 Annotations

-0.68

eatures

-0.65

levant

-0.64

Loading

-0.64

 Definitely

-0.63

POSITIVE LOGITS

 tolerate

1.06

 excuses

1.02

 hesitate

0.97

 impunity

0.95

 relent

0.94

 surrender

0.94

 succumb

0.93

 compl

0.92

partisan

0.92

 compromise

0.91

Activations Density 0.391%

strong statements about standing up for justice, accountability, and equality

negative phrases indicating resistance or refusal toward actions and behaviors

No Comments

No Known Activations