INDEX

Explanations

words related to death or negative outcomes

oai_token-act-pair · gpt-3.5-turbo

occurrences of the prefix "mor" in various 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

racted

-0.68

BOOK

-0.68

 Prohibition

-0.66

ACTED

-0.65

rike

-0.64

ivity

-0.63

htaking

-0.62

Published

-0.61

Minimum

-0.61

payer

-0.61

POSITIVE LOGITS

mor

1.18

gue

0.86

aceutical

0.84

gage

0.80

iday

0.79

phe

0.77

usalem

0.75

osure

0.74

itary

0.72

ilateral

0.72

Activations Density 0.007%

words related to death or negative outcomes

occurrences of the prefix "mor" in various contexts

No Comments

No Known Activations