INDEX

Explanations

capital letters followed by numbers, often formatted like product codes or identifiers

oai_token-act-pair · gpt-3.5-turbo

specific brands or product identifiers associated with various industries

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

taboola

-0.78

ItemTracker

-0.67

 conspicuous

-0.67

alam

-0.62

 bour

-0.60

 Beng

-0.59

scl

-0.59

licted

-0.59

 glow

-0.58

 mere

-0.58

POSITIVE LOGITS

ATA

1.11

senal

1.08

OPS

1.07

OSS

1.06

AR

1.05

RA

1.05

TO

1.02

ELS

1.00

ARE

1.00

OP

1.00

Activations Density 0.150%

capital letters followed by numbers, often formatted like product codes or identifiers

specific brands or product identifiers associated with various industries

No Comments

No Known Activations