INDEX

Explanations

words related to spatial locations or geographical features

oai_token-act-pair · gpt-3.5-turbo

non-standard or unusual characters or symbols

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 spoiler

-0.67

Interested

-0.60

RIP

-0.60

istration

-0.60

 Paulo

-0.60

 Customs

-0.59

 cheers

-0.59

MSN

-0.59

RAFT

-0.59

 Fitzpatrick

-0.57

POSITIVE LOGITS

¿½

0.73

¶æ

0.70

ktop

0.68

cknow

0.68

 tremend

0.68

 glutamate

0.67

·

0.65

Ĥª

0.65

irez

0.65

Īè

0.64

Activations Density 0.000%

words related to spatial locations or geographical features

non-standard or unusual characters or symbols

No Comments

No Known Activations