INDEX

Explanations

phrases indicating uniqueness or first occurrences of events or concepts

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

0:0.01

1:0.06

2:0.24

3:0.06

4:0.01

5:0.03

6:0.07

7:0.07

8:0.10

9:0.09

10:0.10

11:0.10

Negative Logits

ija

-1.27

Tool

-1.05

itivity

-1.02

owship

-0.97

HQ

-0.96

Match

-0.95

onomy

-0.94

cade

-0.92

ampions

-0.92

Sharp

-0.92

POSITIVE LOGITS

 anymore

1.75

nor

1.62

liest

1.38

lier

1.10

 wedd

1.09

iest

1.06

nor

1.06

 Byron

1.06

 ones

1.05

 Browne

0.97

Activations Density 0.107%

phrases indicating uniqueness or first occurrences of events or concepts

No Comments

No Known Activations