INDEX

Explanations

phrases containing the word "proving."

oai_token-act-pair · gpt-3.5-turbo

gerunds, particularly those related to actions or processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

å¯

-0.69

 diapers

-0.64

 controlled

-0.64

 residue

-0.63

 electing

-0.61

 pooled

-0.60

 redistributed

-0.60

inj

-0.60

 identifying

-0.59

 certific

-0.59

POSITIVE LOGITS

ving

1.10

ãĥ¼ãĥĨãĤ£

0.92

tons

0.82

rue

0.82

ves

0.81

HAM

0.81

gence

0.80

theless

0.79

ricanes

0.79

lass

0.78

Activations Density 0.007%

phrases containing the word "proving."

gerunds, particularly those related to actions or processes

No Comments

No Known Activations