INDEX

Explanations

terms related to museums and science

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

dn

-0.60

 stew

-0.60

 unfamiliar

-0.60

 fest

-0.59

 dumps

-0.59

 reins

-0.58

 inward

-0.57

 sworn

-0.56

 undertaking

-0.55

 blitz

-0.55

POSITIVE LOGITS

ļéĨĴ

0.92

 Anon

0.82

MSN

0.81

 Logged

0.81

SEE

0.81

 ÃĹ

0.79

 Photo

0.77

@@

0.76

CAP

0.76

 ..............

0.76

Activations Density 0.120%

terms related to museums and science

No Comments

No Known Activations