INDEX

Explanations

words related to specific locations or places

oai_token-act-pair · gpt-3.5-turbo

terms associated with artistic talent or creative environments

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

 Weasley

-0.60

 angels

-0.59

verb

-0.58

 towels

-0.57

 escorted

-0.56

-0.55

 crawled

-0.55

 Pelosi

-0.55

lin

-0.55

 Dion

-0.54

POSITIVE LOGITS

llah

0.78

 mast

0.76

mast

0.76

assis

0.74

omi

0.73

hesis

0.73

rophe

0.71

riage

0.69

oko

0.67

oji

0.67

Activations Density 0.094%

words related to specific locations or places

terms associated with artistic talent or creative environments

No Comments

No Known Activations