INDEX

Explanations

people's names or pronouns referring to specific individuals

oai_token-act-pair · gpt-3.5-turbo

references to individuals or entities defined by the pronoun "whom."

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

trap

-0.70

termin

-0.69

artifacts

-0.65

repre

-0.64

starting

-0.64

-0.60

 inhibitor

-0.59

hig

-0.59

 pend

-0.59

pillar

-0.59

POSITIVE LOGITS

soever

2.13

she

0.91

he

0.90

we

0.89

 they

0.85

 thou

0.80

 critics

0.80

 Vanity

0.77

you

0.74

 Chomsky

0.72

Activations Density 0.026%

people's names or pronouns referring to specific individuals

references to individuals or entities defined by the pronoun "whom."

No Comments

No Known Activations