INDEX

Explanations

proper nouns related to historical or literary figures

oai_token-act-pair · gpt-3.5-turbo

names of historical figures, particularly those from ancient Rome and related 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

arten

-0.94

lag

-0.89

tw

-0.87

rien

-0.83

caster

-0.80

eals

-0.80

need

-0.79

WAY

-0.77

taboola

-0.77

ynthesis

-0.75

POSITIVE LOGITS

 Caesar

1.45

 Augustus

1.18

 Claud

1.15

 Britann

0.99

 Nero

0.98

 Romans

0.93

 amph

0.90

 Gaul

0.87

 Titus

0.86

olini

0.83

Activations Density 0.012%

proper nouns related to historical or literary figures

names of historical figures, particularly those from ancient Rome and related contexts

No Comments

No Known Activations