INDEX

Explanations

words indicating emphasis or importance

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

hou

-0.74

ysis

-0.72

âĢ¦]

-0.72

ipop

-0.72

ously

-0.70

dal

-0.65

iste

-0.65

rette

-0.64

IGH

-0.64

rences

-0.64

POSITIVE LOGITS

_-

0.86

 namely

0.70

yes

0.64

->

0.63

 albeit

0.61

without

0.60

aka

0.60

 Instruments

0.59

especially

0.59

which

0.59

Activations Density 0.048%

words indicating emphasis or importance

No Comments

No Known Activations