INDEX

Explanations

terms related to marking or identification, possibly involving evaluation or categorization

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "marks" which denotes significance or highlights in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ILLE

-0.68

 Garner

-0.68

æ©Ł

-0.68

ADS

-0.68

erest

-0.65

ECD

-0.65

 disproportionate

-0.64

 Fernand

-0.61

onz

-0.59

 Huang

-0.59

POSITIVE LOGITS

manship

1.46

marks

1.07

 marks

1.00

mas

0.94

mark

0.89

terday

0.89

emark

0.89

hift

0.87

pring

0.84

pelled

0.84

Activations Density 0.007%

terms related to marking or identification, possibly involving evaluation or categorization

occurrences of the word "marks" which denotes significance or highlights in various contexts

No Comments

No Known Activations