INDEX

Explanations

instances of the word "MORE" in various contexts, indicating emphasis or quantity

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

Negative Logits

 Laksh

-0.75

 Sear

-0.72

Barn

-0.71

Rust

-0.70

ãĤ¡

-0.68

 Guth

-0.68

 Gates

-0.68

 Hash

-0.65

 Roof

-0.65

ãĥ©

-0.65

POSITIVE LOGITS

opl

0.71

umbnails

0.69

ratulations

0.68

daq

0.67

 partner

0.64

 standpoint

0.62

 cautiously

0.60

 confid

0.59

 correspondent

0.59

 hopeful

0.59

Activations Density 0.038%

instances of the word "MORE" in various contexts, indicating emphasis or quantity

No Known Activations