INDEX

Explanations

references to official documents or legal forms

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

 honorable

-0.65

âĶĢâĶĢâĶĢâĶĢ

-0.62

{:

-0.61

REDACTED

-0.57

)).

-0.56

âĶ

-0.56

ãĥ´

-0.56

anooga

-0.56

\\\\\\\\

-0.55

apter

-0.55

POSITIVE LOGITS

 amid

1.13

 because

0.88

 during

0.84

 fearing

0.83

 whenever

0.83

 despite

0.83

 anytime

0.82

 lest

0.81

 lately

0.81

 unless

0.80

Activations Density 0.590%

references to official documents or legal forms

No Comments

No Known Activations