INDEX

Explanations

medical and security-related terms and concepts

oai_token-act-pair · gpt-3.5-turbo

terms related to medical, political, and military contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

agame

-0.69

 Heller

-0.69

mma

-0.65

iphate

-0.64

undrum

-0.63

 CrossRef

-0.62

etheless

-0.61

arcity

-0.61

 Poor

-0.61

shapeshifter

-0.60

POSITIVE LOGITS

 items

0.92

 equipment

0.92

 activities

0.88

 clients

0.87

goers

0.87

 packages

0.85

 applications

0.85

 detainees

0.83

 components

0.83

 services

0.82

Activations Density 0.639%

medical and security-related terms and concepts

terms related to medical, political, and military contexts

No Comments

No Known Activations