INDEX

Explanations

references to specific entities or objects

oai_token-act-pair · gpt-3.5-turbo

references to polls and surveys

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Helpful

-0.73

 ..........

-0.59

-0.56

 Skyrim

-0.56

ynthesis

-0.56

cale

-0.55

à¼

-0.55

\.

-0.55

 Flavoring

-0.55

Ô

-0.54

POSITIVE LOGITS

 itself

0.76

 trio

0.73

iest

0.71

 portion

0.70

 designation

0.68

 landscape

0.68

liest

0.67

 acronym

0.66

 playbook

0.65

 fray

0.65

Activations Density 0.601%

references to specific entities or objects

references to polls and surveys

No Comments

No Known Activations