INDEX

Explanations

references to unsuccessful attempts or outcomes

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "failed" in relation to various unsuccessful events or attempts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

utra

-0.88

istics

-0.87

ollen

-0.82

cko

-0.80

aws

-0.79

ensor

-0.78

Center

-0.76

Whereas

-0.74

alon

-0.74

anguage

-0.73

POSITIVE LOGITS

 attempts

1.05

 attempt

1.03

 pregnancies

0.87

 unsuccessful

0.86

 pregnancy

0.85

 miser

0.83

Attempts

0.77

 abort

0.77

 completion

0.70

 Attempt

0.70

Activations Density 0.054%

references to unsuccessful attempts or outcomes

occurrences of the word "failed" in relation to various unsuccessful events or attempts

No Comments

No Known Activations