INDEX

Explanations

phrases related to utilizing or accessing resources or opportunities

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 discipl

-0.70

cz

-0.69

 modification

-0.66

adder

-0.64

ateur

-0.62

weather

-0.62

 Boxing

-0.62

itary

-0.61

EngineDebug

-0.60

loss

-0.59

POSITIVE LOGITS

 into

0.97

 INTO

0.93

uries

0.89

 onto

0.84

 Into

0.82

 holes

0.74

out

0.74

onite

0.71

 shoulders

0.70

kered

0.70

Activations Density 0.070%

phrases related to utilizing or accessing resources or opportunities

No Comments

No Known Activations