INDEX

Explanations

negative or unfavorable descriptors

oai_token-act-pair · gpt-3.5-turbo

negative descriptors of individuals or groups

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

inoa

-0.83

HCR

-0.83

ingham

-0.77

inez

-0.77

aver

-0.77

inet

-0.76

ablished

-0.75

inances

-0.74

ucket

-0.74

avia

-0.74

POSITIVE LOGITS

 nasty

1.10

 surprises

1.00

 adolesc

0.93

 little

0.77

mouth

0.77

 spoil

0.74

 smelling

0.74

 poisonous

0.74

 ugly

0.73

 rotten

0.72

Activations Density 0.008%

negative or unfavorable descriptors

negative descriptors of individuals or groups

No Comments

No Known Activations