INDEX

Explanations

instances where the phrase "no" is immediately followed by the number 9

oai_token-act-pair · gpt-3.5-turbo

statements of negation or lack of certainty

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

rex

-0.71

ale

-0.70

ammad

-0.70

leaning

-0.66

south

-0.66

interstitial

-0.61

often

-0.61

everyone

-0.60

tenance

-0.59

 rhet

-0.59

POSITIVE LOGITS

 clue

1.28

 idea

1.07

 shortage

1.04

 illusions

1.04

 doubt

1.04

 intention

1.04

 excuse

1.02

 qual

1.01

 chance

0.98

 recourse

0.97

Activations Density 0.047%

instances where the phrase "no" is immediately followed by the number 9

statements of negation or lack of certainty

No Comments

No Known Activations