INDEX

Explanations

references to financial penalties or punishments

oai_token-act-pair · gpt-3.5-turbo

mentions of financial penalties or fines

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

OPS

-0.71

owers

-0.70

edes

-0.69

leader

-0.68

OB

-0.67

Ops

-0.67

oir

-0.67

avez

-0.66

groups

-0.66

ipel

-0.66

POSITIVE LOGITS

 fines

1.20

 confir

0.92

 fined

0.90

alties

0.88

 levied

0.87

 evasion

0.78

 redes

0.78

 viol

0.78

 penalties

0.74

 undermin

0.71

Activations Density 0.005%

references to financial penalties or punishments

mentions of financial penalties or fines

No Comments

No Known Activations