INDEX

Explanations

words related to specific proper nouns, such as names of companies and individuals

oai_token-act-pair · gpt-3.5-turbo

specific brands, terms, and proper nouns related to media and technology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

fare

-0.81

orders

-0.77

anim

-0.76

soDeliveryDate

-0.76

kered

-0.74

hov

-0.73

aldi

-0.72

ters

-0.71

dfx

-0.71

apers

-0.68

POSITIVE LOGITS

ments

0.77

 Creed

0.67

ircraft

0.66

pport

0.66

perture

0.65

 Catalyst

0.65

 foil

0.64

DEF

0.62

illary

0.60

qua

0.59

Activations Density 0.343%

words related to specific proper nouns, such as names of companies and individuals

specific brands, terms, and proper nouns related to media and technology

No Comments

No Known Activations