INDEX

Explanations

phrases related to conflicts or disagreements

oai_token-act-pair · gpt-3.5-turbo

references to likelihoods or probabilities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

activate

-0.63

LC

-0.63

Gas

-0.62

Produ

-0.61

san

-0.61

unit

-0.60

activated

-0.60

orate

-0.59

cle

-0.59

gment

-0.59

POSITIVE LOGITS

 odds

4.37

 chances

2.16

 probabilities

1.71

Odd

1.63

 likelihood

1.58

 probability

1.49

dds

1.43

 bets

1.33

 chance

1.13

 risks

1.11

Activations Density 0.006%

phrases related to conflicts or disagreements

references to likelihoods or probabilities

No Comments

No Known Activations