INDEX

Explanations

numerical indicators related to parts or sections of documents

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

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L7_Hcat_z_lr1.20e-03_l11.10e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

float32

Hook Name

blocks.7.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.23

1:0.03

2:0.02

3:0.11

4:0.06

5:0.19

6:0.04

7:0.01

8:0.13

9:0.10

10:0.02

11:0.01

Negative Logits

 slightest

-1.94

 wisely

-1.90

soever

-1.81

 heartbeat

-1.80

actively

-1.79

 whatsoever

-1.79

 notwithstanding

-1.73

 wink

-1.73

existent

-1.69

 veins

-1.67

POSITIVE LOGITS

rint

2.07

CRIPTION

1.96

��

1.93

ategory

1.90

 recap

1.83

rought

1.82

allery

1.82

OTO

1.82

 Analysis

1.79

podcast

1.78

Activations Density 0.035%

numerical indicators related to parts or sections of documents

No Comments

No Known Activations