INDEX

Explanations

mentions of scaling in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to scaling actions or concepts within various contexts

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

ido

-0.73

est

-0.70

Liber

-0.70

Il

-0.67

 Republic

-0.65

vim

-0.65

 Newsletter

-0.65

 Reflex

-0.64

 Liber

-0.64

ttes

-0.64

POSITIVE LOGITS

 scaling

3.51

 scaled

2.96

 scal

2.11

Scale

1.66

 Scal

1.63

 scalable

1.61

 scale

1.54

 scales

1.54

 Scale

1.32

cale

1.31

Activations Density 0.027%

mentions of scaling in various contexts

references to scaling actions or concepts within various contexts

No Comments

No Known Activations