INDEX

Explanations

pronouns and possessive pronouns followed by action verbs

oai_token-act-pair · gpt-3.5-turbo

references to individuals and their accomplishments or attributes within a team context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 spoilers

-0.72

govtrack

-0.66

 correcting

-0.65

 offending

-0.65

 Serious

-0.65

 Probe

-0.64

 Crim

-0.63

checking

-0.62

 blaming

-0.61

 apologizing

-0.59

POSITIVE LOGITS

 afforded

1.12

 bestowed

0.95

 fost

0.89

 glean

0.87

izons

0.87

 cultivate

0.85

 accrued

0.83

 gained

0.83

 attain

0.83

eng

0.82

Activations Density 0.151%

pronouns and possessive pronouns followed by action verbs

references to individuals and their accomplishments or attributes within a team context

No Comments

No Known Activations