INDEX

Explanations

words related to following or succeeding in a sequence

oai_token-act-pair · gpt-3.5-turbo

phrases indicating continuity or follow-up events

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

opathic

-0.68

________________________________________________________________

-0.68

osc

-0.64

.;

-0.64

('

-0.62

ãĥİ

-0.61

 entails

-0.60

 forgot

-0.60

 violates

-0.60

 survives

-0.59

POSITIVE LOGITS

Days

0.92

 months

0.91

 another

0.88

 renewed

0.85

 revelations

0.84

 developments

0.82

 weeks

0.82

 broader

0.80

 intens

0.78

 recent

0.76

Activations Density 0.180%

words related to following or succeeding in a sequence

phrases indicating continuity or follow-up events

No Comments

No Known Activations