INDEX

Explanations

suffixes indicating various forms of action or state, particularly related to verbs or nouns

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_L3_Hcat_z_lr1.20e-03_l19.00e-01_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.3.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

0:0.03

1:0.03

2:0.12

3:0.22

4:0.02

5:0.03

6:0.12

7:0.10

8:0.07

9:0.08

10:0.09

11:0.05

Negative Logits

agame

-1.09

 plateau

-1.04

 Glac

-1.04

ordial

-1.02

 inclination

-1.01

edom

-0.95

 embargo

-0.95

erm

-0.93

aturated

-0.93

 closest

-0.93

POSITIVE LOGITS

Jr

1.41

naire

1.04

LLP

1.03

Topic

1.03

 Detention

1.03

 Correctional

1.00

 Execution

0.97

ERSON

0.96

Topics

0.96

ffe

0.96

Activations Density 0.040%

suffixes indicating various forms of action or state, particularly related to verbs or nouns

No Comments

No Known Activations