INDEX

Explanations

references to specific techniques or methodologies

oai_token-act-pair · gpt-3.5-turbo

mentions of techniques

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

riel

-0.70

isle

-0.67

pg

-0.64

endar

-0.63

ippers

-0.62

minster

-0.61

 Barkley

-0.60

present

-0.60

à

-0.59

onen

-0.58

POSITIVE LOGITS

ologies

1.07

 techniques

0.89

sonian

0.79

 technique

0.76

OLOGY

0.75

 chops

0.73

ologically

0.72

ologic

0.72

 methods

0.72

Methods

0.69

Activations Density 0.020%

references to specific techniques or methodologies

mentions of techniques

No Comments

No Known Activations