INDEX

Explanations

numbers or quantities that are estimated

oai_token-act-pair · gpt-3.5-turbo

terms and phrases related to statistical estimates and approximations

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

feeding

-0.88

woods

-0.77

oard

-0.75

matter

-0.74

FACE

-0.70

interrupted

-0.69

gob

-0.69

Interstitial

-0.69

erton

-0.68

perty

-0.68

POSITIVE LOGITS

 estimates

0.98

imate

0.93

 estimate

0.92

 estimating

0.83

imates

0.81

 distances

0.79

 estimated

0.77

 approximate

0.76

 discrepancies

0.75

 probabilities

0.75

Activations Density 0.039%

numbers or quantities that are estimated

terms and phrases related to statistical estimates and approximations

No Comments

No Known Activations