INDEX

Explanations

quantitative metrics and statistics regarding various subjects

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

Synopsis

-0.77

omen

-0.69

plays

-0.62

sonian

-0.61

BUT

-0.60

ESE

-0.60

 Encyclopedia

-0.54

tick

-0.53

quote

-0.53

changes

-0.53

POSITIVE LOGITS

 thereafter

1.13

 respectively

0.97

 apiece

0.89

 afterward

0.79

 elsewhere

0.71

 versa

0.71

oppable

0.70

 increments

0.69

 afterwards

0.68

qus

0.65

Activations Density 0.116%

quantitative metrics and statistics regarding various subjects

No Comments

No Known Activations