INDEX

Explanations

phrases expressing a strong opinion or belief

oai_token-act-pair · gpt-3.5-turbo

statements related to the nature of concepts or ideas

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

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

 sidx

-0.80

stood

-0.78

mentioned

-0.75

tackle

-0.72

 Tackle

-0.70

ilde

-0.65

ESE

-0.63

DF

-0.63

WER

-0.63

sers

-0.63

POSITIVE LOGITS

 somehow

1.60

 inherently

1.17

 merely

1.16

 magically

1.08

 necessarily

1.05

 solely

1.04

 illegitimate

1.03

 irrelevant

1.01

 meaningless

0.98

 intrinsically

0.98

Activations Density 0.254%

phrases expressing a strong opinion or belief

statements related to the nature of concepts or ideas

No Comments

No Known Activations