INDEX

Explanations

specific technical terms related to engineering or design processes

oai_token-act-pair · gpt-3.5-turbo

sequences of blank tokens or empty content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

enegger

-0.98

ONSORED

-0.79

aeda

-0.76

Interstitial

-0.74

ADRA

-0.74

 Vaugh

-0.69

 conclud

-0.67

CLASSIFIED

-0.66

selves

-0.65

lished

-0.65

POSITIVE LOGITS

tones

0.58

coin

0.57

mont

0.55

rop

0.55

 coats

0.54

anc

0.54

 shader

0.53

pool

0.53

bot

0.53

\":

0.53

Activations Density 0.736%

specific technical terms related to engineering or design processes

sequences of blank tokens or empty content

No Comments

No Known Activations