INDEX

Explanations

terms related to lightweight or lightweight objects

oai_token-act-pair · gpt-3.5-turbo

terms related to lightweight concepts or items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

itia

-0.77

sis

-0.75

owitz

-0.73

thus

-0.71

leon

-0.70

ibel

-0.69

GM

-0.69

ICLE

-0.68

ulhu

-0.68

awar

-0.67

POSITIVE LOGITS

 lightweight

0.97

weight

0.87

 minded

0.86

weights

0.84

 heavyweight

0.82

 dipping

0.75

heartedly

0.72

ened

0.71

 lifting

0.70

 stout

0.69

Activations Density 0.006%

terms related to lightweight or lightweight objects

terms related to lightweight concepts or items

No Comments

No Known Activations