INDEX

Explanations

descriptions of criminal charges and accusations

oai_token-act-pair · gpt-3.5-turbo

instances of the word "charged" related to legal contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

COR

-0.86

 Birth

-0.80

birth

-0.79

Wo

-0.73

 livest

-0.70

icz

-0.69

Birth

-0.69

chan

-0.66

quickShipAvailable

-0.66

PsyNetMessage

-0.64

POSITIVE LOGITS

heet

1.09

 charges

0.93

llah

0.88

 charged

0.85

 accused

0.82

eters

0.80

charges

0.79

 accuses

0.77

 indicted

0.77

 accusing

0.77

Activations Density 0.024%

descriptions of criminal charges and accusations

instances of the word "charged" related to legal contexts

No Comments

No Known Activations