INDEX

Explanations

phrases that indicate a response or reaction to questions or events

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_L5_Hcat_z_lr1.20e-03_l11.00e+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.5.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

0:0.02

1:0.01

2:0.07

3:0.12

4:0.10

5:0.04

6:0.05

7:0.25

8:0.05

9:0.04

10:0.13

11:0.06

Negative Logits

lif

-1.51

 specialize

-1.40

 monop

-1.40

 ladder

-1.39

 equivalents

-1.36

pedia

-1.35

PHI

-1.34

 membr

-1.34

ufact

-1.33

 Gutenberg

-1.33

POSITIVE LOGITS

 queries

2.15

 criticisms

2.01

 criticism

2.00

Question

1.91

 questions

1.87

 objections

1.81

 Questions

1.80

 plaint

1.78

 questioning

1.77

 query

1.72

Activations Density 0.003%

phrases that indicate a response or reaction to questions or events

No Comments

No Known Activations