INDEX

Explanations

the word "report."

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_L10_Hcat_z_lr1.20e-03_l11.30e+00_ds24576_bs4096_dc1.00e-05_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.10.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

0:0.08

1:0.08

2:0.09

3:0.08

4:0.07

5:0.07

6:0.08

7:0.07

8:0.08

9:0.08

10:0.08

11:0.08

Negative Logits

�

-3.04

 Approach

-2.98

Cyn

-2.97

 Plat

-2.67

 Psychology

-2.61

�

-2.60

 Clim

-2.59

�

-2.56

 Push

-2.55

CL

-2.53

POSITIVE LOGITS

abad

3.05

osta

2.97

unda

2.67

cakes

2.66

efer

2.66

Spoiler

2.63

 Weinstein

2.63

ringe

2.59

Jones

2.57

hetto

2.56

Activations Density 0.000%

the word "report."

No Comments

No Known Activations