INDEX

Explanations

terms related to the process of induction or being inducted into a program or group

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

yip

-0.84

urger

-0.80

aeus

-0.79

iot

-0.77

hai

-0.76

ridges

-0.75

upload

-0.75

ecake

-0.74

ersed

-0.73

efully

-0.73

POSITIVE LOGITS

 induct

0.86

ovember

0.74

brance

0.73

äºĶ

0.71

 neuroscience

0.67

 commencement

0.66

MJ

0.62

Lab

0.61

 honorable

0.61

 Fourth

0.61

Activations Density 0.182%

terms related to the process of induction or being inducted into a program or group

No Comments

No Known Activations