INDEX

Explanations

celebrities or characters nominated for awards

oai_token-act-pair · gpt-3.5-turbo

terminologies related to supporting roles and nominations in performance categories

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

nown

-0.89

deal

-0.74

killer

-0.73

oufl

-0.73

chemy

-0.72

ping

-0.71

kay

-0.71

Chem

-0.70

ldom

-0.70

azard

-0.70

POSITIVE LOGITS

 Supports

0.95

 Supporting

0.80

 supporting

0.79

 Supporters

0.76

 support

0.73

Supporters

0.73

iary

0.73

 Supported

0.71

 supports

0.70

heses

0.69

Activations Density 0.029%

celebrities or characters nominated for awards

terminologies related to supporting roles and nominations in performance categories

No Comments

No Known Activations