INDEX

Explanations

negations

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

0:0.07

1:0.09

2:0.07

3:0.13

4:0.07

5:0.08

6:0.07

7:0.04

8:0.08

9:0.18

10:0.04

11:0.04

Negative Logits

��

-2.06

 paradox

-2.04

 difficulties

-2.01

 limbo

-1.99

 dreams

-1.97

 incomp

-1.86

�

-1.84

 contradictions

-1.83

 crashes

-1.83

 crashing

-1.82

POSITIVE LOGITS

 thereto

2.15

 THEM

2.11

soever

2.06

CDC

2.01

.","

2.00

 Monitor

1.97

 dstg

1.96

 Introduced

1.95

Serial

1.92

 NETWORK

1.92

Activations Density 0.000%

negations

No Comments

No Known Activations