INDEX

Explanations

occurrences of the word "the."

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

thood

-0.76

Spread

-0.69

amphetamine

-0.69

FIELD

-0.69

illion

-0.68

ÃĥÃĤ

-0.67

adesh

-0.66

illac

-0.65

abel

-0.65

acea

-0.65

POSITIVE LOGITS

 interviewer

1.02

 Associated

0.99

 Huffington

0.99

BBC

0.98

 latter

0.98

 latest

0.97

 Guardian

0.93

 same

0.91

 Chronicle

0.90

ABC

0.87

Activations Density 0.044%

occurrences of the word "the."

No Comments

No Known Activations