INDEX

Explanations

phrases that denote disclaimers or statements of personal views/representations

oai_token-act-pair · gpt-3.5-turbo

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

aston

-0.68

ionage

-0.64

 quieter

-0.59

 silent

-0.59

 amazed

-0.59

 myst

-0.59

 Chall

-0.59

 beaut

-0.59

 surprise

-0.58

Geh

-0.58

POSITIVE LOGITS

official

0.75

nor

0.74

nor

0.74

 ILCS

0.74

äºº

0.70

osher

0.65

QB

0.64

å¤©

0.63

 views

0.61

Council

0.61

Activations Density 0.100%

phrases that denote disclaimers or statements of personal views/representations

No Comments

No Known Activations