INDEX

Explanations

references to transcripts or quoting of spoken interactions

oai_token-act-pair · gpt-3.5-turbo

references to transcripts and documentation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

ikes

-0.66

Cru

-0.66

 Virtue

-0.66

uu

-0.66

oodoo

-0.62

 Kali

-0.61

sticks

-0.61

ju

-0.61

bike

-0.61

pots

-0.60

POSITIVE LOGITS

 transcript

4.08

 transcripts

3.46

 Transcript

3.03

 transcription

2.24

 transc

1.52

 excerpts

1.36

 translation

1.24

 excerpt

1.21

 mRNA

1.16

 translator

1.10

Activations Density 0.014%

references to transcripts or quoting of spoken interactions

references to transcripts and documentation

No Comments

No Known Activations