INDEX

Explanations

terms related to technology, sci-fi, and technical specifications

oai_token-act-pair · gpt-3.5-turbo

names and titles of various types of media and products

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 tremend

-0.70

 deduct

-0.57

 allowances

-0.53

 soph

-0.53

 selective

-0.52

 offending

-0.52

ependent

-0.51

blogs

-0.51

 dissatisfaction

-0.51

 patience

-0.50

POSITIVE LOGITS

 âĵĺ

0.69

eus

0.68

Ltd

0.67

ombat

0.65

III

0.64

ichi

0.63

ensis

0.61

ÃŃn

0.60

III

0.60

ator

0.59

Activations Density 1.123%

terms related to technology, sci-fi, and technical specifications

names and titles of various types of media and products

No Comments

No Known Activations