INDEX

Explanations

phrases related to time, specifically years and time periods

oai_token-act-pair · gpt-3.5-turbo

references to yearly occurrences or processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

SourceFile

-0.82

anooga

-0.76

 domains

-0.70

onyms

-0.68

hent

-0.67

utf

-0.66

ongh

-0.66

uyomi

-0.65

atche

-0.65

Els

-0.64

POSITIVE LOGITS

 reckoning

0.83

 Celebration

0.71

EVA

0.69

 kickoff

0.69

 touring

0.69

redits

0.66

 onwards

0.65

 celebration

0.64

 spent

0.63

 probation

0.61

Activations Density 0.314%

phrases related to time, specifically years and time periods

references to yearly occurrences or processes

No Comments

No Known Activations