INDEX

Explanations

references to the need for something significant or important

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_post_32k-oai

Configuration

jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

32,768

Data Type

torch.float32

Hook Name

blocks.8.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

ardless

-0.70

fox

-0.70

orest

-0.68

metry

-0.68

 Mahm

-0.68

RANT

-0.68

alde

-0.68

abal

-0.67

arte

-0.66

nih

-0.65

POSITIVE LOGITS

 refres

1.06

lease

0.84

 rethink

0.79

 scapego

0.77

 explanation

0.77

 excuse

0.77

 Citation

0.74

 reminder

0.74

 backup

0.72

 holidays

0.71

Activations Density 0.445%

references to the need for something significant or important

No Comments

No Known Activations