INDEX

Explanations

locations or addresses

oai_token-act-pair · gpt-3.5-turbo

references to locations and organizations

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 realise

-0.81

 realised

-0.75

 analyse

-0.74

 learnt

-0.72

Firstly

-0.71

avour

-0.69

Secondly

-0.69

 organise

-0.67

 Whilst

-0.66

 analys

-0.66

POSITIVE LOGITS

 POLITICO

0.64

 statewide

0.64

 DeVos

0.61

 Tampa

0.61

GOP

0.61

 Asheville

0.60

 Wisconsin

0.59

 Wyoming

0.59

 Denver

0.59

 Arkansas

0.58

Activations Density 3.573%

locations or addresses

references to locations and organizations

No Comments

No Known Activations