INDEX

Explanations

tech-related terms, possibly related to hardware specifications or product features

oai_token-act-pair · gpt-3.5-turbo

numerical or list-like sequences and annotations in a document

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 conclud

-0.83

 millenn

-0.83

 volunte

-0.75

emort

-0.74

packed

-0.73

ertodd

-0.73

 footing

-0.73

 stimulus

-0.72

kefeller

-0.72

 livest

-0.69

POSITIVE LOGITS

 Lanc

0.88

 Dominican

0.78

 Saga

0.78

 Galile

0.77

 Duchess

0.77

 Symphony

0.76

Viz

0.76

 Accord

0.76

 Wand

0.75

 Corpus

0.74

Activations Density 0.555%

tech-related terms, possibly related to hardware specifications or product features

numerical or list-like sequences and annotations in a document

No Comments

No Known Activations