INDEX

Explanations

assertions or opinions repeated multiple times

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.03

1:0.12

2:0.04

3:0.04

4:0.01

5:0.07

6:0.08

7:0.05

8:0.05

9:0.06

10:0.34

11:0.04

Negative Logits

 mounts

-2.00

 Masquerade

-1.93

 bundles

-1.83

ibrary

-1.83

 deliveries

-1.81

 Harmony

-1.80

 cram

-1.75

 housed

-1.74

 undert

-1.74

neys

-1.74

POSITIVE LOGITS

think

2.75

 think

2.69

 thinks

2.44

Think

2.28

 Defin

2.25

thought

2.23

 thought

2.22

 Thinking

2.16

 Think

2.13

isive

2.02

Activations Density 0.029%

assertions or opinions repeated multiple times

No Comments

No Known Activations