INDEX

Explanations

terms related to data processing and algorithms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res_post_32k-oai

Configuration

jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_10.pt

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

32,768

Data Type

torch.float32

Hook Name

blocks.10.hook_resid_post

Hook Layer

Architecture

standard

Context Size

Dataset

Skylion007/openwebtext

Activation Function

topk

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Kard

-0.85

 Canad

-0.79

 Chero

-0.79

 Buckingham

-0.74

 helium

-0.71

 unlaw

-0.69

Ott

-0.68

 Celt

-0.67

Gol

-0.66

cha

-0.66

POSITIVE LOGITS

ushing

0.79

anqu

0.77

aning

0.74

inion

0.73

ESE

0.72

ength

0.70

ocument

0.69

arted

0.69

zee

0.69

ework

0.68

Activations Density 0.155%

terms related to data processing and algorithms

No Comments

No Known Activations