INDEX

Explanations

instances of the word "outlined" and other words indicating structure or planning

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res_post_32k-oai

Configuration

jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_6.pt

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

32,768

Data Type

torch.float32

Hook Name

blocks.6.hook_resid_post

Hook Layer

Architecture

standard

Context Size

Dataset

Skylion007/openwebtext

Activation Function

topk

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Islands

-0.73

-0.67

 Animals

-0.62

etheless

-0.62

 Cage

-0.61

 Money

-0.60

 Ferdinand

-0.59

oro

-0.58

 CLSID

-0.57

ridges

-0.57

POSITIVE LOGITS

ï¸ı

0.81

iflower

0.71

eous

0.70

halla

0.70

umen

0.67

witz

0.67

butt

0.66

 GOODMAN

0.63

advertisement

0.63

peed

0.63

Activations Density 0.231%

instances of the word "outlined" and other words indicating structure or planning

No Comments

No Known Activations