INDEX

Explanations

movie titles

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

interstitial

-0.71

 reconc

-0.66

hunt

-0.64

:[

-0.64

azine

-0.63

selves

-0.63

gress

-0.62

uggest

-0.61

,—

-0.61

POSITIVE LOGITS

?)

0.95

 supra

0.80

huh

0.78

etc

0.78

LLC

0.77

Jr

0.76

Inc

0.76

Pt

0.67

eh

0.67

et

0.66

Activations Density 0.110%

movie titles

No Comments

No Known Activations