INDEX

Explanations

names of vehicles such as brands/types of cars

oai_token-act-pair · gpt-3.5-turbo

specific abbreviations or acronyms, particularly those with 'V' or 'G'

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Creed

-0.58

Zac

-0.58

CPR

-0.57

bun

-0.57

 requisite

-0.57

 commissions

-0.57

 Takeru

-0.56

Suz

-0.55

 lest

-0.55

 frame

-0.54

POSITIVE LOGITS

bilt

1.08

emort

0.99

namese

0.98

amins

0.89

Leaks

0.88

iane

0.86

²¾

0.84

quez

0.83

oslav

0.79

 versa

0.79

Activations Density 0.188%

names of vehicles such as brands/types of cars

specific abbreviations or acronyms, particularly those with 'V' or 'G'

No Comments

No Known Activations