INDEX

Explanations

values or scales expressed as numeric ranges

oai_token-act-pair · gpt-3.5-turbo

rating systems and numerical evaluations

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

 redist

-0.55

 Bounty

-0.54

acas

-0.50

tug

-0.49

¬¼

-0.48

 organizing

-0.48

 Flavoring

-0.47

 crowd

-0.47

 spat

-0.46

 histories

-0.46

POSITIVE LOGITS

neutral

0.73

renheit

0.70

Unknown

0.67

+.

0.67

NULL

0.66

0.65

-.

0.65

-$

0.65

cellent

0.65

 minus

0.64

Activations Density 0.863%

values or scales expressed as numeric ranges

rating systems and numerical evaluations

No Comments

No Known Activations