INDEX

Explanations

terms related to specific names or entities

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names of people or places

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 âĢº

-0.75

 ......

-0.68

¶ħ

-0.59

+++

-0.57

 ..........

-0.56

çĶ°

-0.56

ä»£

-0.55

æĸ¹

-0.54

hement

-0.54

vg

-0.53

POSITIVE LOGITS

gate

0.55

owan

0.55

aney

0.54

elia

0.54

igan

0.54

avia

0.53

reth

0.53

mun

0.52

fen

0.52

fal

0.52

Activations Density 0.669%

terms related to specific names or entities

proper nouns, particularly names of people or places

No Comments

No Known Activations