INDEX

Explanations

quantitative and comparative phrases, particularly related to levels, targets, and standards

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

awa

-0.64

 suspic

-0.62

 differe

-0.61

versive

-0.59

orkshire

-0.58

 incorpor

-0.58

âĹ¼

-0.57

OUP

-0.57

 indist

-0.57

 redes

-0.57

POSITIVE LOGITS

of

0.86

 threshold

0.86

 achievable

0.84

(>

0.81

 allowable

0.81

(~

0.81

 estimate

0.79

 requirement

0.79

($

0.78

 recommended

0.78

Activations Density 0.284%

quantitative and comparative phrases, particularly related to levels, targets, and standards

No Comments

No Known Activations