INDEX

Explanations

entities related to people and organizational structures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-att_32k-oai

Configuration

jbloom/GPT2-Small-OAI-v5-32k-attn-out-SAEs/v5_32k_layer_3

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

32,768

Data Type

torch.float32

Hook Name

blocks.3.hook_attn_out

Hook Layer

Architecture

standard

Context Size

Dataset

Skylion007/openwebtext

Activation Function

topk

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ĪĴ

-0.71

utics

-0.68

 Malf

-0.67

 Transformers

-0.66

istries

-0.64

uffs

-0.63

ws

-0.62

 Mell

-0.61

HC

-0.61

raf

-0.59

POSITIVE LOGITS

 ende

0.74

gol

0.71

omsky

0.69

 Semin

0.65

osuke

0.65

 offic

0.63

 shooter

0.62

 Hilbert

0.62

 COUR

0.61

ategor

0.60

Activations Density 0.052%

entities related to people and organizational structures

No Comments

No Known Activations