INDEX

Explanations

information related to something causing negative feelings or unease

oai_token-act-pair · gpt-3.5-turbo

instances of the word "disturbing."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

cellence

-0.90

thood

-0.77

á

-0.76

haps

-0.74

eva

-0.74

ophers

-0.73

oned

-0.72

ript

-0.72

eah

-0.71

ardi

-0.71

POSITIVE LOGITS

 undermin

0.80

 disturbing

0.75

 sexist

0.73

 noises

0.73

 disturb

0.73

ingly

0.72

 alarms

0.70

 creeps

0.69

ly

0.65

 Cree

0.65

Activations Density 0.014%

information related to something causing negative feelings or unease

instances of the word "disturbing."

No Comments

No Known Activations