INDEX

Explanations

prefixes related to negation such as "de-" or "-ing"

oai_token-act-pair · gpt-3.5-turbo

terms related to deconstruction or decentralization

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

LW

-0.76

Ys

-0.69

 revel

-0.66

 sorely

-0.64

 regards

-0.61

 upwards

-0.61

 outstanding

-0.60

 sincerely

-0.58

 defeats

-0.58

LV

-0.58

POSITIVE LOGITS

emphasis

1.23

fact

1.20

escal

1.17

commit

1.07

construct

1.04

radical

1.03

syn

1.02

conf

1.00

context

1.00

human

0.99

Activations Density 0.020%

prefixes related to negation such as "de-" or "-ing"

terms related to deconstruction or decentralization

No Comments

No Known Activations