INDEX

Explanations

words related to preventing negative events or outcomes

oai_token-act-pair · gpt-3.5-turbo

phrases that emphasize prevention

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ammy

-0.85

night

-0.70

edded

-0.70

eah

-0.69

ingers

-0.62

elt

-0.62

bard

-0.60

geist

-0.60

Truth

-0.60

MON

-0.60

POSITIVE LOGITS

ative

1.01

 detection

0.84

 inhib

0.83

ively

0.80

 regress

0.79

 duplicate

0.78

ministic

0.73

auga

0.70

 accidental

0.68

 obstruct

0.68

Activations Density 0.031%

words related to preventing negative events or outcomes

phrases that emphasize prevention

No Comments

No Known Activations