INDEX

Explanations

pronouns or possessive determiners used to refer to a specific person or group

oai_token-act-pair · gpt-3.5-turbo

possessive pronouns and expressions of ownership or reference

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

lished

-0.83

-+-+

-0.82

tu

-0.79

alde

-0.78

pps

-0.74

Lago

-0.73

ellen

-0.72

oak

-0.71

llah

-0.70

nell

-0.69

POSITIVE LOGITS

 cues

1.31

cue

1.28

 chances

1.06

 rightful

1.06

 oath

0.95

 plunge

0.93

 frustrations

0.91

own

0.89

 bearings

0.88

 pulse

0.87

Activations Density 0.034%

pronouns or possessive determiners used to refer to a specific person or group

possessive pronouns and expressions of ownership or reference

No Comments

No Known Activations