INDEX

Explanations

technical or coded references in structured data or programming 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

staking

-0.68

 Speedway

-0.66

 pall

-0.64

worth

-0.63

neys

-0.61

kers

-0.59

uls

-0.58

 Katrina

-0.56

urdue

-0.56

anooga

-0.56

POSITIVE LOGITS

ouble

0.93

yss

0.88

isl

0.82

oÄŁan

0.81

party

0.80

hin

0.77

ividual

0.77

iamond

0.73

sg

0.72

irty

0.72

Activations Density 0.010%

technical or coded references in structured data or programming contexts

No Comments

No Known Activations