INDEX

Explanations

adjectives related to qualities or characteristics

oai_token-act-pair · gpt-3.5-turbo

terms related to the validity and significance of evidence or claims made

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

cellent

-0.79

alian

-0.79

ometimes

-0.74

mostly

-0.73

essler

-0.71

sometimes

-0.71

 Mostly

-0.70

 Initially

-0.69

last

-0.69

ften

-0.68

POSITIVE LOGITS

 whatsoever

1.35

nor

1.15

 repercussions

1.14

 reperc

1.09

 explanation

1.08

 affiliation

1.03

 indication

1.02

 reason

1.00

 recourse

1.00

 distinction

0.98

Activations Density 0.187%

adjectives related to qualities or characteristics

terms related to the validity and significance of evidence or claims made

No Comments

No Known Activations