INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

items or events associated with significant individuals or roles in society

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

ELF

-0.61

ãĥ¥

-0.61

(>

-0.59

ESA

-0.59

ize

-0.58

Different

-0.56

gress

-0.55

ORT

-0.55

ente

-0.55

TOR

-0.53

POSITIVE LOGITS

 wrote

0.96

 joins

0.95

 remembers

0.91

 teaches

0.90

 joined

0.90

 became

0.89

 testified

0.88

 took

0.88

 reacted

0.87

 flew

0.85

Activations Density 0.167%

proper nouns

items or events associated with significant individuals or roles in society

No Comments

No Known Activations