INDEX

Explanations

technical terms and names related to politics and finance

oai_token-act-pair · gpt-3.5-turbo

instances of surprise or unexpected reactions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Extend

-0.59

 ILCS

-0.56

-,

-0.55

Interstitial

-0.54

 hinge

-0.54

)/

-0.53

 breaker

-0.52

âĢ²

-0.52

 detriment

-0.52

 fade

-0.52

POSITIVE LOGITS

published

0.57

 Canaver

0.56

 plagiar

0.54

radical

0.54

Pai

0.54

 Priebus

0.52

 prophetic

0.52

ullivan

0.52

UCS

0.51

 scathing

0.51

Activations Density 0.971%

technical terms and names related to politics and finance

instances of surprise or unexpected reactions

No Comments

No Known Activations