INDEX

Explanations

temperature-related keywords

oai_token-act-pair · gpt-3.5-turbo

numerical data and comparisons related to levels and thresholds

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Psy

-0.76

UST

-0.75

actionDate

-0.71

IGHTS

-0.70

åĪ

-0.68

ographics

-0.66

ãĥĪ

-0.64

fortune

-0.63

 Cosponsors

-0.63

Ô

-0.62

POSITIVE LOGITS

 achievable

1.08

 allowable

1.03

 increments

0.93

 cutoff

0.84

(>

0.81

 corresponds

0.78

 attained

0.77

 approximation

0.76

 achieved

0.75

imum

0.75

Activations Density 0.382%

temperature-related keywords

numerical data and comparisons related to levels and thresholds

No Comments

No Known Activations