INDEX

Explanations

prefixes in technical contexts

oai_token-act-pair · gpt-3.5-turbo

occurrences of prefixes and suffixes in technical contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

neau

-0.84

ONSORED

-0.73

zeb

-0.72

APH

-0.71

Hoy

-0.70

dfx

-0.70

ndra

-0.70

 Weir

-0.69

 Quinn

-0.69

 Stephenson

-0.68

POSITIVE LOGITS

 prefix

1.28

 suffix

1.06

prefix

0.97

es

0.94

entimes

0.90

ãĥ¼ãĥĨãĤ£

0.88

 masks

0.82

forth

0.79

pse

0.79

 mask

0.79

Activations Density 0.018%

prefixes in technical contexts

occurrences of prefixes and suffixes in technical contexts

No Comments

No Known Activations