INDEX

Explanations

locations or distances

oai_token-act-pair · gpt-3.5-turbo

distances or locations marked by "away"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

oly

-0.85

conn

-0.67

cess

-0.67

 streng

-0.63

nce

-0.62

alg

-0.62

Ku

-0.62

erity

-0.61

 Metatron

-0.60

gram

-0.59

POSITIVE LOGITS

ruck

0.79

lier

0.76

 away

0.70

 Away

0.66

aneously

0.66

away

0.64

 WHEN

0.62

 Jagu

0.62

heid

0.62

Registered

0.61

Activations Density 0.017%

locations or distances

distances or locations marked by "away"

No Comments

No Known Activations