INDEX

Explanations

adjectives related to stability

oai_token-act-pair · gpt-3.5-turbo

terms associated with stability

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

 appealed

-0.76

FactoryReloaded

-0.73

sbm

-0.72

inx

-0.69

hig

-0.69

oths

-0.68

 Outs

-0.66

orig

-0.66

retty

-0.65

NS

-0.64

POSITIVE LOGITS

mates

1.06

mate

1.04

 isot

1.00

iable

0.86

ament

0.82

 equilibrium

0.80

 Stability

0.75

 enough

0.74

 stable

0.73

ative

0.72

Activations Density 0.016%

adjectives related to stability

terms associated with stability

No Comments

No Known Activations