INDEX

Explanations

words related to a specific historical era, particularly the Obama-era

oai_token-act-pair · gpt-3.5-turbo

references to specific historical periods or eras, 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 @ 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

olulu

-0.83

sie

-0.76

doi

-0.70

apego

-0.69

hematic

-0.68

etsk

-0.66

 mathemat

-0.64

lay

-0.63

alam

-0.62

irlf

-0.61

POSITIVE LOGITS

era

1.15

ERA

0.92

itton

0.85

interstitial

0.75

osaurs

0.73

isal

0.73

ffiti

0.72

JA

0.70

velength

0.69

Era

0.69

Activations Density 0.007%

words related to a specific historical era, particularly the Obama-era

references to specific historical periods or eras, particularly related to political contexts

No Comments

No Known Activations