INDEX

Explanations

actions and events described in legal or crime-related contexts

oai_token-act-pair · gpt-3.5-turbo

punctuation marks and their placement in sentences

oai_token-act-pair · gpt-4o-mini Triggered by @bot

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

ãĤ´ãĥ³

-0.90

eah

-0.78

¥µ

-0.77

vre

-0.77

ãĤ¦ãĤ¹

-0.74

Probably

-0.71

elta

-0.70

enz

-0.69

ãĥ¥

-0.68

apest

-0.68

POSITIVE LOGITS

 however

1.26

he

0.89

she

0.88

 though

0.86

upon

0.82

 Sergeant

0.80

 they

0.80

 meanwhile

0.76

 surrounded

0.75

 another

0.74

Activations Density 0.190%

actions and events described in legal or crime-related contexts

punctuation marks and their placement in sentences

No Comments

No Known Activations