INDEX

Explanations

phrases that indicate value or quality, particularly related to improvement or satisfaction

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

2:0.17

3:0.08

4:0.01

5:0.06

6:0.11

7:0.10

8:0.08

9:0.13

10:0.08

11:0.06

Negative Logits

feld

-1.28

oway

-1.06

visor

-1.06

acons

-1.06

acon

-1.04

ican

-1.01

iq

-1.01

overed

-1.01

kell

-1.01

atis

-1.00

POSITIVE LOGITS

 fundamentals

1.11

 fulfil

0.92

 neurotrans

0.92

 stripes

0.91

 gratification

0.91

 fundamental

0.90

 residues

0.89

IPM

0.89

 consciousness

0.88

 dstg

0.88

Activations Density 0.137%

phrases that indicate value or quality, particularly related to improvement or satisfaction

No Comments

No Known Activations