INDEX

Explanations

words related to location, particularly referencing the city "Toronto."

oai_token-act-pair · gpt-3.5-turbo

repeated occurrences of the word "on" and other location or action indicators

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Camb

-0.68

 fault

-0.67

 Virgin

-0.65

MSM

-0.63

 Verse

-0.61

 Canaan

-0.60

 surv

-0.59

 illeg

-0.59

 discriminating

-0.57

LIA

-0.57

POSITIVE LOGITS

cot

0.86

ulum

0.85

oots

0.83

hiba

0.82

hing

0.80

ango

0.79

opot

0.79

oos

0.77

ule

0.76

edo

0.76

Activations Density 0.098%

words related to location, particularly referencing the city "Toronto."

repeated occurrences of the word "on" and other location or action indicators

No Comments

No Known Activations