INDEX

Explanations

number sequences typically representing statistics or scores in sports contexts

oai_token-act-pair · gpt-3.5-turbo

negative indicators or occurrences in sports statistics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 Dickinson

-0.86

ulhu

-0.84

oleon

-0.75

achus

-0.70

 Twain

-0.69

 Seym

-0.66

 wedd

-0.64

 Affect

-0.63

yssey

-0.62

 Compass

-0.62

POSITIVE LOGITS

degree

1.05

sized

1.03

based

0.98

level

0.94

hour

0.90

plus

0.87

ranked

0.87

bound

0.84

year

0.83

inch

0.83

Activations Density 0.276%

number sequences typically representing statistics or scores in sports contexts

negative indicators or occurrences in sports statistics

No Comments

No Known Activations