INDEX

Explanations

proper nouns, specifically names of individuals and positions 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 @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

 Rebellion

-0.74

 Rebels

-0.73

 Wonderland

-0.69

 Cerberus

-0.68

 Leviathan

-0.68

 theaters

-0.66

 robbers

-0.65

 Colleges

-0.65

 womb

-0.64

 machines

-0.64

POSITIVE LOGITS

ergus

1.01

imir

0.96

frey

0.96

ileen

0.95

jit

0.95

ureen

0.94

iji

0.93

anya

0.93

resa

0.93

jan

0.91

Activations Density 0.174%

proper nouns, specifically names of individuals and positions related to political contexts

No Comments

No Known Activations