INDEX

Explanations

proper nouns, specifically names of individuals

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals, particularly related to political 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

enegger

-0.90

oster

-0.76

alore

-0.73

ilde

-0.73

redit

-0.72

aber

-0.71

anooga

-0.71

inances

-0.70

fleet

-0.70

rarily

-0.69

POSITIVE LOGITS

 Sonia

0.93

 Gandhi

0.87

 Abedin

0.74

cles

0.73

OTUS

0.72

 Tome

0.72

Nad

0.69

éĥ

0.68

 Rural

0.65

ata

0.64

Activations Density 0.013%

proper nouns, specifically names of individuals

references to specific individuals, particularly related to political contexts

No Comments

No Known Activations