INDEX

Explanations

technical terms and details in a text

oai_token-act-pair · gpt-3.5-turbo

references to methodologies and evaluative judgments about them

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 adolescence

-0.57

 grooming

-0.52

aila

-0.50

 puberty

-0.49

 sentencing

-0.49

 corros

-0.49

ursday

-0.47

 hazards

-0.47

 Sabbath

-0.46

 disturbances

-0.46

POSITIVE LOGITS

;)

0.92

+.

0.90

_.

0.90

 altogether

0.90

:)

0.89

*.

0.85

 itself

0.84

:-)

0.84

().

0.82

 ðŁĻĤ

0.82

Activations Density 0.893%

technical terms and details in a text

references to methodologies and evaluative judgments about them

No Comments

No Known Activations