INDEX

Explanations

terms related to manageable tasks or situations

oai_token-act-pair · gpt-3.5-turbo

terms related to manageability and recognizability

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

oly

-0.84

ocrates

-0.82

orius

-0.80

oult

-0.80

olog

-0.79

otom

-0.78

cair

-0.78

otomy

-0.78

inen

-0.77

erva

-0.76

POSITIVE LOGITS

 Haram

0.80

isable

0.80

 readable

0.77

izable

0.75

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

0.74

 behavi

0.70

ãĥķ

0.70

lihood

0.70

 Jarvis

0.69

 behav

0.69

Activations Density 0.024%

terms related to manageable tasks or situations

terms related to manageability and recognizability

No Comments

No Known Activations