INDEX

Explanations

terminology related to references or additional information

oai_token-act-pair · gpt-3.5-turbo

occurrences of the phrase "see also," indicating references or additional information

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.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

liest

-0.72

athi

-0.67

UGE

-0.65

agus

-0.63

science

-0.62

 Derby

-0.61

odan

-0.60

elong

-0.60

etime

-0.60

adoes

-0.59

POSITIVE LOGITS

 includes

0.80

 referred

0.76

roph

0.75

 known

0.74

 included

0.73

 Accessed

0.72

 spelled

0.70

 optionally

0.69

 noteworthy

0.68

 occasionally

0.67

Activations Density 0.090%

terminology related to references or additional information

occurrences of the phrase "see also," indicating references or additional information

No Comments

No Known Activations