INDEX

Explanations

names or terms related to locations or people with the string "oa" or "OA"

oai_token-act-pair · gpt-3.5-turbo

references to specific geographical locations or regions

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

parts

-0.80

stroke

-0.79

netic

-0.76

sonian

-0.76

advertisement

-0.74

Catal

-0.74

tick

-0.73

nces

-0.71

 glim

-0.69

stem

-0.68

POSITIVE LOGITS

uth

1.28

uthor

1.12

velength

0.98

issance

0.98

pport

0.93

zzo

0.91

oa

0.91

verage

0.86

ignt

0.84

ppo

0.83

Activations Density 0.007%

names or terms related to locations or people with the string "oa" or "OA"

references to specific geographical locations or regions

No Comments

No Known Activations