INDEX

Explanations

references to fragments and fragmentation

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

Demand

-0.80

Keefe

-0.73

OPA

-0.70

orney

-0.68

rollers

-0.66

rol

-0.66

ILLE

-0.65

reon

-0.65

LOS

-0.65

IAS

-0.63

POSITIVE LOGITS

 fragments

1.15

 fragment

0.95

glass

0.91

mentation

0.91

mented

0.87

 Frag

0.85

 fragmented

0.83

atis

0.80

ions

0.79

bite

0.79

Activations Density 0.008%

references to fragments and fragmentation

No Comments

No Known Activations