INDEX

Explanations

references to rewards or payments for capturing someone or providing information

oai_token-act-pair · gpt-3.5-turbo

mentions of bounties and related concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

icken

-0.79

andem

-0.76

ppard

-0.74

case

-0.74

ROR

-0.70

CON

-0.68

VW

-0.67

 Ange

-0.67

opath

-0.66

orsche

-0.66

POSITIVE LOGITS

 bounty

1.61

 bount

1.42

 Bounty

1.15

 hunter

1.06

 hunters

0.95

iless

0.93

vertisement

0.84

 reward

0.82

ounty

0.78

 rewards

0.77

Activations Density 0.008%

references to rewards or payments for capturing someone or providing information

mentions of bounties and related concepts

No Comments

No Known Activations